Optimal traffic safety alternative index selection method and system suitable for complex road nodes

Abstract

The invention discloses an optimal traffic safety alternative index selection method and system suitable for a complex road node. The method comprises the steps of delimiting a road region boundary of the complex road node and storing the boundary in a database in a coordinate mode; complete microscopic trajectory data of all vehicles in the nodes are periodically obtained, and data preprocessing and parameter extraction are carried out; carrying out node dangerous driving scene judgment and verification to obtain dangerous scene data, and carrying out state division; constructing a candidate replacement index set, and calculating to obtain each candidate replacement index value for the dangerous scene; constructing an evaluation index set of the alternative indexes, and calculating to obtain evaluation index values for the candidate alternative indexes; and calculating the distance between each candidate substitution index and the corresponding optimal scheme and the distance between each candidate substitution index and the corresponding worst scheme, and selecting the traffic safety substitution index with the highest relative proximity to the optimal scheme as the optimal traffic safety substitution index. According to the method, the characteristics of each index are fully considered, and the limitation of a single index is avoided.

Description

Technical Field

[0001] The present invention relates to the field of traffic safety alternative index selection, and in particular to an optimal traffic safety alternative index selection method and system suitable for complex road nodes. Background Art

[0002] For a long time, traffic risk assessment has generally used traffic accident statistics to evaluate road traffic safety. However, traditional traffic accident evaluation methods are difficult to implement. With the gradual maturity of traffic conflict theory, traffic safety alternative indicators have gradually replaced traditional accident statistics with their advantages of large samples, small cycles, small areas, and high reliability, and have been widely used in safety evaluation and prediction. At present, due to the limitations of the applicable scenarios of a single traffic conflict alternative indicator, many self-made indicators and composite indicators have gradually been used in traffic safety assessment.

[0003] At present, there is no definite standard for the selection of safety alternative indicators. Especially for complex road nodes such as toll station diversion areas, the selection of different safety alternative indicators will have a great impact on the accuracy of safety assessment results. Therefore, it is necessary to make reasonable selections before safety evaluation. In addition, traditional safety evaluation indicators have their own limitations and inconsistencies, and it is difficult to obtain accurate prediction results in practical applications. Therefore, the use of composite improved metrics has better universality than single indicators, but there is a lack of evaluation systems that compare composite indicators with traditional single indicators. In view of this, it is crucial to propose a more systematic alternative traffic safety evaluation scheme in order to improve the accuracy of traffic safety evaluation results. Summary of the invention

[0004] In view of the problem that there is no standard selection method for safety alternative indicators for complex road nodes and that existing studies mostly use a single indicator to evaluate traffic safety, the present invention provides a method and system for selecting optimal traffic safety alternative indicators suitable for complex road nodes. The method compares traffic conflict alternative indicators horizontally under specific dangerous scenarios, which can overcome the problem that traditional traffic accident evaluation methods are difficult to handle samples and data, and achieves the goal of improving the accuracy of traffic safety evaluation results.

[0005] In a first aspect, the present invention provides a method for selecting an optimal traffic safety alternative indicator applicable to a complex road node, comprising:

[0006] S1: Delineate the boundaries of the road area of ​​complex road nodes and store them in the database in the form of coordinates;

[0007] S2: Periodically obtain the microscopic trajectory data of all vehicles in the complex road nodes, perform data preprocessing to obtain vehicle trajectory data samples, and further extract parameters;

[0008] S3: Based on the vehicle trajectory data samples and extracted parameters, the dangerous driving scenarios at complex road nodes are determined and verified to obtain dangerous scenario data;

[0009] S4: Construct a set of candidate traffic safety alternative indicators, calculate the values ​​of each candidate traffic safety alternative indicator for dangerous scenarios, and divide them into low-risk and high-risk states;

[0010] S5: construct an evaluation index set of traffic safety alternative indicators, and calculate and obtain the value of each evaluation index for each candidate traffic safety alternative indicator, wherein the evaluation index includes accuracy, rapidity, and severity;

[0011] S06: According to the values ​​of each evaluation index, the distance between each candidate traffic safety alternative index and the corresponding optimal solution and the worst solution is calculated, and the traffic safety alternative index with the highest relative proximity to the optimal solution is selected as the optimal traffic safety alternative index.

[0012] Furthermore, the specific process of S1 is:

[0013] S11. Delineation of road area boundaries at complex road nodes: Identify complex road nodes, such as intersections, multi-level overpasses, roundabouts, etc., through high-precision map data and satellite images; select multiple representative points in the identified area, which include key locations such as road center lines, curb lines, and traffic island edges; based on the selected representative points, apply boundary fitting algorithms, such as convex hull algorithms or polygon fitting algorithms, to generate boundary polygons of the road area to accurately describe the geometric shape and range of complex road nodes.

[0014] S12. Establish a database based on the acquired coordinates of the road area boundaries of complex road nodes: perform coordinate conversion on the coordinates of each vertex of the generated boundary polygon, that is, convert from the map coordinate system to the global positioning system (GPS) coordinate system or other applicable geographic coordinate systems to ensure coordinate consistency between different data sources; store the converted boundary coordinate data in a structured form in the database, wherein the database design adopts a relational database or a spatial database, and the specific table structure includes fields such as node ID, boundary vertex coordinates (longitude and latitude), and timestamp; the storage process covers data insertion, updating, and backup to ensure data integrity and security.

[0015] S13. For the built database, in order to ensure the accuracy and completeness of the boundary data in the database, the boundary data stored in the database shall be verified, including comparison with field measurement data or verification through field investigation. Based on the verification results, inaccurate or incomplete boundary data shall be adjusted and corrected to ensure that it truly reflects the actual situation of complex road nodes.

[0016] Furthermore, step S2 includes:

[0017]

[0017] [S21. The acquisition of vehicle microscopic trajectory data for complex road nodes is to collect vehicle movement trajectory data for the entire area of complex road nodes by using roadside perception devices (such as millimeter-wave radars and lidar) or aerial photography, and perform data preprocessing, where the data preprocessing includes:

[0018] Data cleaning: Remove invalid values, error values, and duplicate values from the data;

[0019] Data synchronization: Synchronize the data obtained by different sensors in terms of time to ensure data consistency;

[0020] Time alignment: Align the data collected at different time points in a time series to ensure data consistency on the time axis;

[0021] Spatial correction: For the data obtained by aerial photography or roadside perception devices, spatial correction is required to convert the image data into actual ground coordinates to ensure data accuracy;

[0022]

[0022] [S22. Process the vehicle trajectory data samples through roadside edge computing devices to extract parameters and obtain vehicle feature parameters and environmental feature parameters; among them, the vehicle feature parameters include: vehicle number, position coordinates, heading angle, speed, acceleration, vehicle type, lane where the vehicle is located, distance, time, and the environmental feature parameters include the distance to the vehicle in front and the speed of the vehicle in front.

[0023] Preferably, the acquisition of vehicle microscopic trajectory data for complex road nodes is also carried out by taking videos of the survey location, using video automatic recognition technology to separate the background from the video background, and tracking the vehicle to obtain continuous vehicle microscopic trajectory data.

[0024] Furthermore, the step S3 includes:

[0025] The determination of the dangerous driving scenario is based on the vehicle feature parameters (such as vehicle speed, distance, time, position coordinates) extracted from the preprocessed vehicle trajectory data samples, and discriminates the head center point of each vehicle C f at each moment t from the instantaneous monitoring distance d between the head center point of the vehicle and the tail center point of the vehicle C l in front, and the size relationship between the instantaneous monitoring distance d and the braking distance S of the vehicle itself. When d < S, it is determined that the start time of the dangerous driving scenario of vehicle C f has started When d is less than or equal to 0.5 meters, it is discriminated that a collision has occurred and the end time of the dangerous scenario A total of n dangerous driving scenario data are obtained, and this dangerous scenario data is the reference set A of traffic safety alternative indicators.

[0026] Among them, d is the instantaneous monitoring distance, that is, vehicle Cf The spatial distance between the center point of the vehicle head and the preceding vehicle C l and the center point of the vehicle tail; S is the braking distance of vehicle C f , and its calculation formula is as follows:

[0027]

[0028] In the formula, t1 is the time from when the driver starts to step on the brake pedal until the brake starts to take effect; t2 is the time when the braking force increases; v0 is the initial vehicle braking speed; j max is the maximum braking deceleration.

[0029] And according to the magnitude of the instantaneous monitoring distance d between the center point of the vehicle head of each vehicle C f and the center point of the vehicle tail of the preceding vehicle C at each moment t, all dangerous scenarios are divided into two states: low-risk state and high-risk state: l Low-risk state, that is, each vehicle C

[0030] is in the state of X ≤ d < S, where X is the safety braking distance of the following vehicle C f when the preceding vehicle C is driving normally, and its calculation formula is as follows: l f l v1 is the speed of the preceding vehicle C

[0031]

[0032]

[0033]

[0034] High-risk state, that is, each vehicle C f is in the state of 0 ≤ d < X.

[0035] Furthermore, step S4 includes:

[0036] S41. Construct a set of candidate traffic safety alternative indicators, where the candidate traffic safety alternative indicators include: Time to Collision (TTC), Modified Time to Collision (MTTC), Extended Time to Collision (ETTC), Time to Collision with Driver Reaction (TTCD), Deceleration Rate to Avoid Collision (DRAC), Composite Index COM ettc ;

[0037] S42. For the reference set A of dangerous scenarios, calculate the specific values of each candidate traffic safety alternative indicator for each frame;

[0038] S43. Divide each candidate traffic safety alternative indicator into states, and the states include low-risk state and high-risk state: when the value of each indicator is within the threshold range of the preset low-risk state, it is divided into the low-risk state; when the value of each indicator is within the threshold range of the preset high-risk state, it is divided into the high-risk state;

[0038] Among them, the composite index COM ettc It is based on a vehicle in a certain period of time in the scene T = [t0-t * , t0+t * ] within the mean ETTC mean , extreme ETTC min and the instantaneous ETTC value at time t0 Constructed; where t0 is any time in the dangerous scene reference set A, t * The length of the time interval set manually, COM ettc The value of is calculated once per frame;

[0039] Calculate the instantaneous ETTC for each frame t t value, t is any time, and the instantaneous ETTC t Value preprocessing: When ETTC t <0s or ETTC t >10s, set ETTC t =10s,ETTC mean is the vehicle C in time period T f ETTC at each time t t The mean of ETTC min is the vehicle C in time period T f ETTC t The minimum value of is the instantaneous ETTC value at time t0; ETTC mean ETTC min The calculation formula is as follows:

[0040]

[0041] Furthermore, based on three basic indicators ETTC mean and ETTC min Construction of composite indicator COM ettc The specific process is:

[0042] S421: divide all time periods T into [t0-t * , t0+t * ] to standardize the three basic indicators

[0043]

[0044] Where Y ij is the standardized value of the jth basic indicator in the i-th period; x ijis the value of the jth basic indicator in the i-th period, i = 1, 2, ..., n; j = 1, 2, 3, corresponding to the three basic indicators respectively ETTC mean and ETTC min ;

[0045] S422. Based on the standardized value Y ij , calculate three basic indicators ETTC mean and ETTC min The information entropy is calculated as follows:

[0046]

[0047]

[0048] In the formula, E j is the information entropy of the jth basic indicator; f ij Y is the proportion of the i-th period under the j-th basic indicator to the sum of all sample values ​​of the indicator; ij is the standardized value of the jth basic indicator in the i-th period, i = 1, 2, ..., n; j = 1, 2, 3, corresponding to the three basic indicators respectively ETTC mean and ETTC min ;

[0049] S423. Based on three basic indicators ETTC mean and ETTC min Information entropy E j , calculate the difference coefficient and entropy weight of the three basic indicators:

[0050] G j =1-E j

[0051]

[0052] W={W1,W2,W3}

[0053] Where W j is the entropy weight under the jth basic index; G j is the difference coefficient of the jth basic index, j = 1, 2, 3, corresponding to the three basic indicators respectively ETTC mean and ETTC min ; W is the entropy weight matrix of the three basic indicators;

[0054] S424. Based on the standardized value Y ij and entropy weight W j, construct a weighted decision matrix:

[0055] Z ij =Y ij ×W j

[0056]

[0057] Among them, Z is the weighted decision matrix; Y ij is the standardized value of the jth basic indicator in the i-th period, Z ij W is the weighted normalized value of the jth basic indicator in the ith period; j is the entropy weight under the jth basic index; i = 1, 2, ..., n; j = 1, 2, 3, corresponding to the three basic indicators respectively ETTC mean and ETTC min ;

[0058] S425. Based on the calculated weighted decision matrix Z, obtain the time period with the least conflict and the time period with the greatest conflict among all time periods; wherein the time period with the least conflict refers to the maximum value of each basic indicator in the weighted decision matrix; the time period with the greatest conflict refers to the minimum value of each evaluation indicator in the weighted decision matrix:

[0059]

[0060] Among them, Z + The period with the least conflict; Z - This is the time when the conflict is greatest;

[0061] S426. Calculate the distance between each time period and the time period with the least conflict and the time period with the greatest conflict under the jth basic index:

[0062]

[0063] in, is the distance between the i-th time period and the time period with the least conflict; is the distance between the i-th time period and the time period with the largest conflict; is the period with the least conflict under the jth basic indicator; Z ij It is the normalized value after weighting under the jth basic indicator in the ith period; is the period with the largest conflict under the jth basic indicator; i = 1, 2, ..., n; j = 1, 2, 3, corresponding to the three basic indicators respectively ETTC mean and ETTC min ;

[0064] S427. Based on the distance between each time period and the time period with the least conflict and the time period with the greatest conflict Calculate the relative proximity S between the i-th continuous conflict period and the period with the least conflict i , that is, the composite index of the central time t0 of the i-th period Value, where the relative closeness S i The calculation formula is as follows:

[0065]

[0066] in, is the distance between the i-th time period and the time period with the largest conflict; is the distance between the i-th time period and the time period with the least conflict; i = 1, 2, ..., n; S i is the relative proximity between the i-th time period and the time period with the least conflict. i <1 means the scene is in low-risk state, if 0 i <0.4, the scene is in a high-risk state.

[0067] Furthermore, the evaluation index set of traffic safety alternative indicators refers to the set of evaluation indicators for evaluating the performance of all traffic safety alternative indicators, where the evaluation indicators include accuracy, rapidity, and severity:

[0068] Accuracy refers to the accuracy of each traffic conflict alternative indicator in identifying dangerous scenes. The higher the accuracy, the better. Its calculation refers to the ratio of the number of scenes in the selected reference set A where the current traffic conflict alternative indicator judges each scene as low-risk or high-risk to the number of all dangerous scenes. Taking TTC as an example, for any dangerous scene in the reference set A, the number of times TTC judges the scene as low-risk or high-risk at each moment accounts for more than 50% of the scene duration, which means that the scene is accurately judged as dangerous. For n scenes in the reference set A, there are y scenes with TTC as the judgment indicator. i The accuracy of this indicator is

[0069] Rapidity refers to the rapidity of each traffic conflict alternative index in identifying dangerous or high-risk scenes. The earlier the identification, the better. Its calculation refers to the calculation of any dangerous scene in the reference set A, starting at the time of the i-th scene. To determine the standard time, each traffic conflict index x is used to judge the dangerous state of the same scene. The time when each index first determines the i-th scene as dangerous or high-risk is Get each indicator The time and the i-th scene The time difference between the moments When Δt​i,x When it is positive, it means that the indicator judges the dangerous state of the scene earlier than the manual judgment in the reference set A. At this time, the rapidity of the judgment of each indicator can be expressed as and is a positive value in the range [0,1], where It is the fastest indicator to determine the accuracy among all indicators. The time and the i-th scene The time difference between the moments. i,x When it is a negative value, it means that the indicator’s judgment of the dangerous state of the scene is later than the manual judgment in the reference set A. At this time, the rapidity of the judgment of each indicator can be expressed as in is the end time of the i-th scene, and is a negative value, ranging from (0 to -∞). Taking TTC as an example, in n scenarios, its speed is

[0070] Severity refers to the accuracy of each traffic conflict alternative indicator in identifying various dangerous states of dangerous / high-risk scenes. The higher the accuracy, the better. Its calculation refers to the ratio P1 of the number of dangerous scenes judged by the current traffic conflict alternative indicator to the number of all dangerous scenes in the n dangerous scenes in the selected reference set A, and the ratio P2 of the number of high-risk scenes judged by the current traffic conflict alternative indicator to the number of all high-risk scenes. The average of P1 and P2 is selected as the severity evaluation indicator. Taking TTC as an example,

[0071] Furthermore, the specific process of S6 is:

[0072] S61: Standardize the accuracy, rapidity, and severity values ​​of all candidate traffic safety alternative indicators:

[0073]

[0074] Among them, Y ij is the standardized value of the i-th safety alternative indicator under the j-th evaluation indicator; x ij is the value of the i-th safety alternative index under the j-th evaluation index; i is the safety alternative index, i = 1, 2, ..., n; j is the evaluation index, when j = 1 it is accuracy, when j = 2 it is rapidity, when j = 3 it is severity;

[0075] S62: The standardized value Y of each safety alternative indicator under each evaluation indicator ij , calculate the information entropy E of each evaluation index j and the difference coefficient G j , the formula is as follows:

[0076]

[0077] G j =1-E j

[0078] In the formula, E j is the information entropy of the jth evaluation index, when j = 1 it is accuracy, when j = 2 it is rapidity, when j = 3 it is severity; f ij Y is the proportion of the i-th safety alternative indicator under the j-th evaluation indicator to all sample values ​​of the indicator, i = 1, 2, ..., n; ij is the standardized value of the i-th safety alternative indicator under the j-th evaluation indicator; G j is the difference coefficient of the jth evaluation index;

[0079] S63: Based on the obtained difference coefficient of the j-th evaluation index, calculate and obtain the entropy weight W corresponding to the j-th evaluation index j And establish the entropy weight matrix W:

[0080]

[0081] W={W1,W2,W3}

[0082] Among them, W j is the entropy weight under the jth evaluation index, when j = 1 it is accuracy, when j = 2 it is rapidity, when j = 3 it is severity; G j is the difference coefficient of the jth evaluation index; W is the entropy weight matrix.

[0083] S64: Based on the normalized value Y ij and entropy weight W j , construct a weighted decision matrix:

[0084] Z ij =Y ij ×W j

[0085]

[0086] Among them, Z is the weighted decision matrix; Y ij is the standardized value of the i-th safety alternative indicator under the j-th evaluation indicator, i = 1, 2, ..., n, when j = 1, it is accuracy, when j = 2, it is rapidity, when j = 3, it is severity; W j is the entropy weight under the jth evaluation index;

[0087] S65: The maximum value of each evaluation index in the weighted decision matrix Z is defined as the optimal solution; the minimum value of each evaluation index in the weighted decision matrix Z is defined as the worst solution:

[0088]

[0089] Among them, Z + is the optimal solution; Z - is the worst solution;

[0090] S66: The optimal solution Z based on each evaluation indicator + and the worst solution Z - , calculate the distance between each candidate traffic safety alternative index and the optimal solution and the worst solution under the jth evaluation index:

[0091]

[0092] in, is the distance between the ith traffic safety alternative index and the optimal solution; is the distance between the ith traffic safety alternative index and the optimal solution; is the optimal solution under the jth evaluation index; Z ij is the weighted value of the i-th safety alternative indicator under the j-th evaluation indicator; is the worst solution under the jth evaluation index, i = 1, 2, ..., n, when j = 1 it is accuracy, when j = 2 it is speed, when j = 3 it is severity;

[0093] S67: Based on the distance between each traffic safety alternative index and the optimal and worst solutions Calculate the relative closeness S between the i-th safety alternative index and the optimal solution i , where the relative proximity S i The calculation formula is as follows:

[0094]

[0095] Among them, S i is the relative closeness between the ith traffic safety alternative index and the optimal solution; is the distance between the ith traffic safety alternative index and the worst solution; is the distance between the ith traffic safety alternative index and the optimal solution, i = 1, 2, ..., n,

[0096] S68: Select relative proximity S i The highest safety alternative is the most optimal traffic safety alternative.

[0097] In a second aspect, the present invention provides an optimal traffic safety alternative indicator selection system applicable to complex road nodes, comprising:

[0098] Data acquisition and preprocessing module: used to define the boundaries of road areas at complex road nodes and store them in the database in the form of coordinates; relying on roadside sensing equipment, it periodically acquires the complete microscopic trajectory data of all vehicles in complex road nodes, and performs data preprocessing and parameter extraction;

[0099] Dangerous scene data calculation and acquisition module: used to determine and verify dangerous driving scenes at complex road nodes based on vehicle trajectory data samples and extracted parameters, and obtain dangerous scene data;

[0100] Candidate traffic safety alternative index value acquisition module: used to construct a set of candidate traffic safety alternative indexes, and calculate and obtain the value of each candidate traffic safety alternative index for dangerous scenarios;

[0101] The module for obtaining the optimal traffic safety alternative indicator is used to construct a set of evaluation indicators for traffic safety alternative indicators. For each candidate traffic safety alternative indicator, the module calculates and obtains the value of each evaluation indicator, where the evaluation indicators include accuracy, rapidity, and severity. According to the value of each evaluation indicator, the module calculates the distance between each candidate traffic safety alternative indicator and the corresponding optimal solution and the worst solution, and selects the traffic safety alternative indicator with the highest relative proximity to the optimal solution as the optimal traffic safety alternative indicator.

[0102] Beneficial Effects

[0103] The present invention proposes a method and system for selecting the optimal traffic safety alternative indicator suitable for complex road nodes. The method is aimed at horizontally comparing traffic conflict alternative indicators in complex road nodes, which can overcome the problem of difficulty in processing samples and data in traditional traffic accident evaluation methods, fully consider the characteristics of each indicator, avoid the limitations of a single indicator, and improve the accuracy of traffic safety evaluation results. BRIEF DESCRIPTION OF THE DRAWINGS

[0104] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0105] Figure 1 It is a flow chart of a method for selecting an optimal traffic safety alternative indicator applicable to a complex road node provided by an embodiment of the present invention;

[0106] Figure 2 It is a time distribution diagram of dangerous scenarios provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0107] To make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be described in detail below. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other implementation methods obtained by ordinary technicians in this field without creative work belong to the scope of protection of the present invention.

[0108] Example 1

[0109] like Figure 1 As shown, this embodiment provides a method and system for selecting an optimal traffic safety alternative indicator applicable to complex road nodes, including:

[0110] S1: Delineate the boundaries of the road area at complex road nodes and store them in the database in the form of coordinates.

[0111] S11. Delineation of road area boundaries at complex road nodes: Identify complex road nodes (such as intersections, multi-level overpasses, roundabouts, etc.) through high-precision map data and satellite images; select a number of representative points in the identified area using multi-point sampling methods (such as random sampling, uniform sampling, systematic sampling, etc.) (representative points include key locations such as road center lines, curb lines, and traffic island edges); based on the selected representative points, apply boundary fitting algorithms (such as convex hull algorithms or polygon fitting algorithms) to generate boundary polygons of the road area to accurately describe the geometric shape and range of complex road nodes;

[0112] S12. Establish a database based on the acquired complex road node road area boundary coordinates: perform coordinate conversion on the coordinates of each vertex of the generated boundary polygon, that is, convert from the map coordinate system to the global positioning system (GPS) coordinate system or other applicable geographic coordinate system to ensure the coordinate consistency between different data sources; store the converted boundary coordinate data in a structured form in the database, wherein the database design adopts a relational database or a spatial database, and the specific table structure includes fields such as node ID, boundary vertex coordinates (longitude and latitude), and timestamp; the storage process covers data insertion, update, and backup to ensure data integrity and security;

[0113] S13. For the built database, in order to ensure the accuracy and completeness of the boundary data in the database, the boundary data stored in the database shall be verified, including comparison with field measurement data or verification through field investigation. Based on the verification results, inaccurate or incomplete boundary data shall be adjusted and corrected to ensure that it truly reflects the actual situation of complex road nodes.

[0114] S2: Periodically obtain the complete microscopic trajectory data of all vehicles in complex road nodes, perform data preprocessing to obtain vehicle trajectory data samples, and further extract parameters.

[0115] S21. The acquisition of vehicle micro-trajectory data of complex road nodes is to collect vehicle motion trajectory data of the entire area of ​​the complex road node by using roadside sensing equipment or aerial photography (i.e., vehicle micro-trajectory data is obtained after processing vehicle motion trajectory data), and perform data preprocessing, wherein the data preprocessing includes:

[0116] Data cleaning: remove invalid values, error values ​​and duplicate values ​​in the data;

[0117] Data synchronization: synchronize the time of data acquired by different sensors to ensure data consistency;

[0118] Time alignment: Align the time series of data collected at different time points to ensure the consistency of the data on the time axis;

[0119] Spatial correction: For the data obtained by aerial photography or roadside sensing equipment, spatial correction is required to convert the image data into actual ground coordinates to ensure the accuracy of the data;

[0120] S22. Process the vehicle trajectory data through the roadside edge computing device, extract parameters, and obtain vehicle characteristic parameters and environmental characteristic parameters; among which, the vehicle characteristic parameters include: vehicle number, position coordinates, heading angle, speed, acceleration, vehicle type, and lane; the environmental characteristic parameters include the distance to the vehicle in front and the speed of the vehicle in front.

[0121] Preferably, the vehicle micro-trajectory data of complex road nodes is obtained by shooting a video of the survey location, separating the background from the video background using video automatic recognition technology, and tracking the vehicle to obtain continuous vehicle micro-trajectory data.

[0122] A toll station is selected as the complex road node of this example. The basic road section upstream of the toll station has 4 main lanes with a total width of 15 meters; the starting width of the gradual widening area is 15 meters and the ending width is 65 meters. There are 12 toll lanes in the toll station, including 3 ETC toll lanes set inside the toll station, and the rest are MTC channels. The diversion area of ​​the toll station is 360 meters long, including a 60-meter diversion area with lanes and a 300-meter diversion area without lanes.

[0123] In specific implementation, this embodiment uses the high-point video recording method to take videos with a drone. The drone takes videos vertically above the diversion area at a flight altitude of about 200 meters. Roads with a length of nearly 400 meters and their traffic conditions are captured. The clarity of the videos taken by the drone reaches 4K ultra-high definition at 30 frames per second. To obtain more data, video shooting is carried out during the peak period with a large traffic flow. The early peak period with clear weather and gentle breeze is selected for shooting to ensure the video quality. The shooting duration is selected as 3 hours to collect more sample data. In this example, it is from 7:30 to 8:30 in the early peak period on Friday, Saturday, and Sunday.

[0124] The vehicle trajectory automatic recognition system is used to process the videos taken, identify the driving trajectories of vehicles within the specified area, and extract the trajectory data of each frame of the vehicle. After eliminating errors and data verification, complete vehicle trajectory data of 1103 vehicles are obtained.

[0125] The extracted vehicle trajectory data is based on the data of the video pixel coordinate system, and only includes information such as the position coordinates of the vehicle in each frame, the vehicle ID, and the frame number ID. For the sake of simplifying the processing process, one data is retained every three frames. The innermost end of the diversion area is set as the origin (point O) of the coordinate system, and the x-axis and y-axis are respectively set along the horizontal and vertical directions of the road.

[0126] The vehicle trajectory is based on the data of the video pixel coordinate system. In actual processing, after converting the video pixel coordinate data into geodetic coordinate data, the actual displacement between two adjacent frames of the vehicle can be obtained. The time interval between each frame is 0.1s, and then the actual speed and acceleration of the vehicle in each frame can be output for subsequent determination and verification of dangerous scenarios.

[0127] S3: Based on the vehicle trajectory data samples and the extracted parameters, the dangerous driving scenarios at complex road nodes are discriminated to obtain dangerous scenario data. For the processed vehicle trajectory data samples, microscopic motion parameters such as vehicle speed and distance are extracted, and it is discriminated whether the instantaneous monitoring distance d between the center point of the front of vehicle C f and the center point of the rear of the preceding vehicle C l at each moment t is less than the braking distance S of the vehicle itself. When d < S, it is determined that the start time of the dangerous driving scenario of vehicle C f has occurred. When d is less than or equal to 0.5 meters, it is discriminated that a collision has occurred and it is the end time of the dangerous scenario. A total of 100 dangerous driving scenario data are obtained, and this dangerous scenario data is the reference set A of traffic safety alternative indicators.

[0128] Among them, d is the instantaneous monitoring distance, that is, the spatial distance between the center point of the front of vehicle C f and the center point of the rear of the preceding vehicle C l ; S is the braking distance of vehicle C fThe braking distance is calculated as follows:

[0129] S1 = v0t1

[0130]

[0131] Since t2 < 1, t2 2 can be ignored, and the final formula for S can be obtained as:

[0132]

[0133] where t1 is the time from when the driver starts to step on the brake pedal until the brake starts to take effect; t2 is the time for the braking force to increase; t3 is the duration after the braking force increases to the maximum value; t4 is the time from when the vehicle stops until the brake is released. S is the total braking distance; S1, S2, and S3 respectively represent the distances traveled by the vehicle within t1, t2, and t3; v0 is the initial braking speed of the vehicle; j max is the maximum braking deceleration. In this embodiment, all the observed vehicles are cars. According to the braking process formula of the vehicle and the national standard, it is stipulated that the maximum braking deceleration j max of the car is 7.4 m / s 2 . It is stipulated that for small vehicles, the average value of t1 is 0.09 s, and the average value of t2 is 0.06 s. Thus, the discrimination method for the braking distance is shown in the following table:

[0134] Braking Distance Discrimination Method

[0135]

[0136] And according to the size of the instantaneous monitoring distance d between the center point of the front of each vehicle C f and the center point of the rear of the preceding vehicle C l at each moment t, each moment in each dangerous scenario is divided into two states according to the degree of danger: low-risk state and high-risk state:

[0137] Low-risk state, that is, each vehicle C f is in the state of X ≤ d < S, where X is the safe braking distance of the following vehicle C l when the preceding vehicle C f is traveling normally, and its calculation formula is as follows:

[0138]

[0139] v1 is the speed of the preceding vehicle C l .

[0140] High-risk state, that is, each vehicle C f is in the state of 0 ≤ d < X.

[0141] Based on the extracted trajectory data of 100 dangerous scenes, a distribution diagram of the duration of dangerous scenes is drawn, such as Figure 2 shown.

[0142] S04: Construct a set of candidate traffic safety alternative indicators, and calculate and obtain the value of each candidate traffic safety alternative indicator for dangerous scenarios.

[0143] S41, constructing a set of candidate traffic safety alternative indicators, wherein the candidate traffic safety alternative indicators include: time to collision (TTC), improved time to collision (MTTC), extended time to collision (ETTC), interference time to collision (TTCD), deceleration rate to avoid collision (DRAC), composite indicator COM ettc ;

[0144] S42. For the dangerous scene reference set A, calculate and obtain the specific value of each candidate traffic safety alternative indicator for each frame;

[0145] S43. Classify each candidate traffic safety alternative indicator into states, including low-risk state and high-risk state. When the value of each indicator is within the preset low-risk state threshold range, it is classified as a low-risk state; when the value of each indicator is within the preset high-risk state threshold range, it is classified as a high-risk state. The thresholds of the dangerous state and the high-risk state can be adjusted based on the actual situation. In this embodiment, the threshold values ​​are designed as follows:

[0146] Time to collision TTC: When 1s≤TTC<4s, it is defined as a low-risk state; when 0s≤TTC<1s, it is defined as a high-risk state;

[0147] Improved distance-to-collision time MTTC: when 1s≤MTTC<4s, it is defined as a low-risk state; when 0s≤MTTC<1s, it is defined as a high-risk state;

[0148] Extended distance to collision time ETTC: When 1s≤ETTC<4s, it is defined as a low-risk state; when 0s≤ETTC<1s, it is defined as a high-risk state;

[0149] Interference collision time TTCD: when 1s≤TTCD<4s, it is defined as a low-risk state; when 0s≤TTCD<1s, it is defined as a high-risk state;

[0150] Collision avoidance deceleration rate DRAC: 1.5m / s 2 ≤TTCD<3.5m / s 2 It is defined as a low-risk state; when 3.5m / s 2 ≤TTCD, defined as high-risk status;

[0151] Composite Index COM ettc:When 0.4≤COM ettc <1, it is defined as a low-risk state; when 0≤COM ettc When <0.4, it is defined as a high-risk state.

[0152] Among them, the composite index COM ettc It is based on a vehicle in a certain period of time in the scene T = [t0-t * , t0+t * ] within the mean ETTC mean , extreme ETTC min and the instantaneous ETTC value at time t0 (when time t0 is for a certain period T, it is the center time of period T; when it is for dangerous scene A, it is any time in dangerous scene A; dangerous scene A includes multiple periods T) Constructed; where t0 is any time in the dangerous scene reference set A, t * The length of the time interval set manually, COM ettc The value of is calculated once per frame; is the ETTC value at the central time, ETTC t is any time within the time period T.

[0153] Calculate the instantaneous ETTC for each frame t t value, and for instantaneous ETTC t Value preprocessing: When ETTC t <0s or ETTC t >10s, set ETTC t =10s,ETTC mean is the vehicle C in time period T f ETTC at each time t t The mean of ETTC min is the vehicle C in time period T f ETTC t The minimum value of is the instantaneous ETTC value at time t0; ETTC mean ETTC min The calculation formula is as follows:

[0154]

[0155] In this embodiment, the time-to-collision (TTC) is:

[0156]

[0157] x l,t is the center position of the front vehicle at time t (m), x f,tis the center point position of the rear vehicle at time t (m), L l is the length of the preceding vehicle (m), v f,t is the speed of the following vehicle at time t (m / s), v l,t is the speed of the preceding vehicle at time t (m / s).

[0158] Modified Time-to-Collision (MTTC):

[0159]

[0160] Δa t =a f,t -a l,t

[0161] θv t is the speed difference between the two vehicles before and after time t (m / s), x l,t is the center position of the front vehicle at time t (m), x f,t is the center point position of the rear vehicle at time t (m), L l is the length of the preceding vehicle (m), a l,t is the acceleration of the preceding vehicle at time t (m / s 2 ), a f,t is the acceleration of the following vehicle at time t (m / s 2 ).

[0162] Extended Time-to-Collision (ETTC):

[0163]

[0164] D ij is the distance between the centroids of the two vehicles before and after time t (m), L l is the length of the preceding vehicle (m), L f is the length of the following vehicle (m), O i , O j is the vector representation of the center point (centroid) of the two vehicles, V i 、V j is the vector representation of the speed of the two vehicles.

[0165] Time to Collision with Disturbance (TTCD):

[0166]

[0167] v l is the initial velocity of the vehicle in front (m / s), v f is the initial velocity of the rear vehicle (m / s), d is the deceleration of the front vehicle (m / s2 ), L0 is the initial relative distance between the two vehicles, L l is the length of the preceding vehicle (m).

[0168] Acceleration-based traffic conflict alternative indicator: Deceleration Rate to Avoid Crash (DRAC):

[0169]

[0170] v f,t is the speed of the following vehicle at time t (m / s), v l,t is the speed of the preceding vehicle at time t (m / s), x l,t is the center position of the front vehicle at time t (m), x f,t is the center point position of the rear vehicle at time t (m), L l is the length of the preceding vehicle (m).

[0171] Based on three basic indicators ETT mean and ETTC min Construction of composite indicator COM ettc The specific process is:

[0172] S421: divide all time periods T into [t0-t * , t0+t * ] to standardize the three basic indicators

[0173]

[0174] Where Y ij is the standardized value of the jth basic indicator in the i-th period; x ij is the value of the jth basic indicator in the i-th period, i = 1, 2, ..., n; j = 1, 2, 3, corresponding to the three basic indicators respectively ETTC mean and ETTC min ;

[0175] S422. Based on the standardized value Y ij , calculate three basic indicators ETTC mean and ETTC min The information entropy is calculated as follows:

[0176]

[0177] In the formula, E j is the information entropy of the jth basic indicator; f ijY is the proportion of the i-th period under the j-th basic indicator to the sum of all sample values ​​of the indicator; ij is the standardized value of the jth basic indicator in the i-th period, i = 1, 2, ..., n; j = 1, 2, 3, corresponding to the three basic indicators respectively ETTC mean and ETTC min ;

[0178] S423. Based on three basic indicators ETTC mean and ETTC min Information entropy E j , calculate the difference coefficient and entropy weight of the three basic indicators:

[0179] G j =1-E j

[0180]

[0181] W={W1,W2,W3}

[0182] Where W j is the entropy weight under the jth basic index; G j is the difference coefficient of the jth basic index, j = 1, 2, 3, corresponding to the three basic indicators respectively ETTC mean and ETTC min ; W is the entropy weight matrix of the three basic indicators;

[0183] S424. Based on the standardized value Y ij and entropy weight W j , construct a weighted decision matrix:

[0184] Z ij =Y ij ×W j

[0185]

[0186] Among them, Z is the weighted decision matrix; Y ij is the standardized value of the jth basic indicator in the i-th period, Z ij W is the weighted normalized value of the jth basic indicator in the ith period; j is the entropy weight under the jth basic index; i = 1, 2, ..., n; j = 1, 2, 3, corresponding to the three basic indicators respectively ETTC mean and ETTC min ;

[0187] S425. Based on the calculated weighted decision matrix Z, obtain the time period with the least conflict and the time period with the greatest conflict in all persistent conflict intervals; wherein the time period with the least conflict refers to the maximum value of each basic indicator in the weighted decision matrix; the time period with the greatest conflict refers to the minimum value of each evaluation indicator in the weighted decision matrix:

[0188]

[0189]

[0190] Among them, Z + The period with the least conflict; Z - This is the time when the conflict is greatest;

[0191] S426. Calculate the distance between each time period and the time period with the least conflict and the time period with the greatest conflict under the jth basic index:

[0192]

[0193] in, is the distance between the i-th time period and the time period with the least conflict; is the distance between the i-th time period and the time period with the largest conflict; is the period with the least conflict under the jth basic indicator; Z ij It is the normalized value after weighting under the jth basic indicator in the ith period; is the period with the largest conflict under the jth basic indicator; i = 1, 2, ..., n; j = 1, 2, 3, corresponding to the three basic indicators respectively ETTC mean and ETTC min ;

[0194] S427. Based on the distance between each time period and the time period with the least conflict and the time period with the greatest conflict Calculate the relative proximity S between the i-th continuous conflict period and the period with the least conflict i , that is, the composite index of the central time t0 of the i-th period Value, where the relative closeness S i The calculation formula is as follows:

[0195]

[0196] in, is the distance between the i-th time period and the time period with the largest conflict; is the distance between the i-th time period and the time period with the least conflict; i = 1, 2, ..., n; S i is the relative proximity between the i-th time period and the time period with the least conflict. i<1 means the scene is in low-risk state, if 0 i <0.4, the scene is in a high-risk state.

[0197] S05: Construct an evaluation index set of traffic safety alternative indicators, and calculate and obtain the value of each evaluation index for each candidate traffic safety alternative indicator, wherein the evaluation index includes accuracy, rapidity, and severity.

[0198] The evaluation index set of traffic safety alternative indicators refers to the set of evaluation indicators for evaluating the performance of all traffic safety alternative indicators, where the evaluation indicators include accuracy, rapidity, and severity:

[0199] Accuracy refers to the accuracy of each traffic conflict alternative indicator in identifying dangerous scenes. The higher the accuracy, the better. Its calculation refers to the ratio of the number of scenes in the selected reference set A where the current traffic conflict alternative indicator judges each scene as low-risk or high-risk to the number of all dangerous scenes. Taking TTC as an example, for any dangerous scene in the reference set A, the number of times TTC judges the scene as low-risk or high-risk at each moment accounts for more than 50% of the scene duration, which means that the scene is accurately judged as dangerous. For n scenes in the reference set A, there are y scenes with TTC as the judgment indicator. i The accuracy of this indicator is

[0200] Rapidity refers to the rapidity of each traffic conflict alternative index in identifying dangerous or high-risk scenes. The earlier the identification, the better. Its calculation refers to the calculation of any dangerous scene in the reference set A, starting at the time of the i-th scene. To determine the standard time, each traffic conflict alternative index x is used to judge the dangerous state of the same scene. The time when each index first determines the i-th scene as dangerous or high-risk is Get each indicator The time and the i-th scene The time difference between the moments When Δt i,x When it is positive, it means that the indicator judges the dangerous state of the scene earlier than the manual judgment in the reference set A. At this time, the rapidity of the judgment of each indicator can be expressed as and is a positive value in the range [0,1], where It is the fastest indicator to determine the accuracy among all indicators. The time and the i-th scene The time difference between the moments. i,x When it is a negative value, it means that the indicator’s judgment of the dangerous state of the scene is later than the manual judgment in the reference set A. At this time, the rapidity of the judgment of each indicator can be expressed as in​ is the end time of the i-th scene, and is a negative value, ranging from (0 to -∞). Taking TTC as an example, in n scenarios, its speed is

[0201] Severity refers to the accuracy of each traffic conflict alternative indicator in identifying various dangerous states of dangerous / high-risk scenes. The higher the accuracy, the better. Its calculation refers to the ratio P1 of the number of dangerous scenes judged by the current traffic conflict alternative indicator to the number of all dangerous scenes in the n dangerous scenes in the selected reference set A, and the ratio P2 of the number of high-risk scenes judged by the current traffic conflict alternative indicator to the number of all high-risk scenes. The average of P1 and P2 is selected as the severity evaluation indicator. Taking TTC as an example,

[0202] S6: Calculate the distance between each candidate traffic safety alternative indicator and the corresponding optimal solution and the worst solution according to the value of each evaluation indicator, and select the traffic safety alternative indicator with the highest relative proximity to the optimal solution as the optimal traffic safety alternative indicator;

[0203] S61: Standardize the accuracy, rapidity, and severity values ​​of all candidate traffic safety alternative indicators:

[0204]

[0205] Among them, Y ij is the standardized value of the i-th safety alternative indicator under the j-th evaluation indicator; x ij is the value of the i-th safety alternative index under the j-th evaluation index; i is the safety alternative index, i = 1, 2, ..., n; j is the evaluation index, when j = 1 it is accuracy, when j = 2 it is rapidity, when j = 3 it is severity;

[0206] S62: The standardized value Y of each safety alternative indicator under each evaluation indicator ij , calculate the information entropy E of each evaluation index j and the difference coefficient G j , the formula is as follows:

[0207]

[0208] G j =1-E j

[0209] In the formula, E j is the information entropy of the jth evaluation index, when j = 1 it is accuracy, when j = 2 it is rapidity, when j = 3 it is severity; f ijY is the proportion of the i-th safety alternative indicator under the j-th evaluation indicator to all sample values ​​of the indicator, i = 1, 2, ..., n; ij is the standardized value of the i-th safety alternative indicator under the j-th evaluation indicator; G j is the difference coefficient of the jth evaluation index;

[0210] S63: Based on the obtained difference coefficient of the j-th evaluation index, calculate and obtain the entropy weight W corresponding to the j-th evaluation index j And establish the entropy weight matrix W:

[0211]

[0212] W={W1,W2,W3}

[0213] Among them, W j is the entropy weight under the jth evaluation index, when j = 1 it is accuracy, when j = 2 it is rapidity, when j = 3 it is severity; G j is the difference coefficient of the jth evaluation index; W is the entropy weight matrix.

[0214] S64: Based on the normalized value Y ij and entropy weight W j , construct a weighted decision matrix:

[0215] Z ij =Y ij ×W j

[0216]

[0217] Among them, Z is the weighted decision matrix; Y ij is the standardized value of the i-th safety alternative indicator under the j-th evaluation indicator, i = 1, 2, ..., n, when j = 1, it is accuracy, when j = 2, it is rapidity, when j = 3, it is severity; W j is the entropy weight under the jth evaluation index;

[0218] S65: The maximum value of each evaluation index in the weighted decision matrix Z is defined as the optimal solution; the minimum value of each evaluation index in the weighted decision matrix Z is defined as the worst solution:

[0219]

[0220]

[0221] Among them, Z + is the optimal solution; Z - is the worst solution;

[0222] S66: The optimal solution Z based on each evaluation indicator + and the worst solution Z - , calculate the distance between each candidate traffic safety alternative index and the optimal solution and the worst solution under the jth evaluation index:

[0223]

[0224] in, is the distance between the ith traffic safety alternative index and the optimal solution; is the distance between the ith traffic safety alternative index and the optimal solution; is the optimal solution under the jth evaluation index; Z ij is the weighted value of the i-th safety alternative indicator under the j-th evaluation indicator; is the worst solution under the jth evaluation index, i = 1, 2, ..., n, when j = 1 it is accuracy, when j = 2 it is speed, when j = 3 it is severity;

[0225] S67: Based on the distance between each traffic safety alternative index and the optimal and worst solutions Calculate the relative closeness S between the i-th safety alternative index and the optimal solution i , where the relative proximity S i The calculation formula is as follows:

[0226]

[0227] Among them, S i is the relative closeness between the ith traffic safety alternative index and the optimal solution; is the distance between the ith traffic safety alternative index and the worst solution; is the distance between the ith traffic safety alternative index and the optimal solution, i = 1, 2, ..., n,

[0228] S68: Select relative proximity S i The highest safety alternative is the most optimal traffic safety alternative.

[0229] The calculation formula for comparing the relative closeness of each safety alternative indicator to the optimal solution is as follows:

[0230] i max = argmax i＝1,2,3,...,n S i

[0231] Among them, S i is the relative closeness between the ith traffic safety alternative index and the optimal solution, S i The value is always greater than 0 and less than 1, S iThe closer the value is to 1, the better the security assessment effect of this indicator is. i The highest safety alternative index is the optimal traffic safety alternative index.

[0232] In this example, S6>(S1, S2, S3, S4, S5), so the safety alternative indicator corresponding to S6 is the composite indicator COM ettc It is the most suitable alternative evaluation index for traffic safety at complex road nodes.

[0233] The results of the three comparison methods of accuracy, rapidity and severity are shown in the following table.

[0234] Evaluation results table

[0235]

[0236] The results show that the composite index COM ettc The relative closeness to the optimal solution is the highest among the six candidate traffic safety alternative indicators and is generally the most suitable for traffic safety alternative evaluation in toll station diversion areas.

[0237] Example 2

[0238] In this embodiment, the above method can also be deployed in a road safety monitoring and evaluation system, and the specific steps are as follows:

[0239] Step 1. In the integration and deployment stage, the optimal traffic safety alternative indicator selection method is seamlessly integrated into the existing road safety monitoring and evaluation system.

[0240] First, through modular design and interface definition, smooth data transmission between different functional modules within the system can be achieved.

[0241] Next, a data processing pipeline is established to achieve real-time transmission and processing of traffic data, ensuring that the system responds immediately to road safety conditions.

[0242] Subsequently, unit testing, integration testing, and performance testing are performed to ensure that the modules work together and remain stable under high data volume and high concurrency conditions. Finally, the system is deployed to an optimized server environment for trial operation and real-time monitoring to ensure the stability and reliability of the system.

[0243] Step 2. In the continuous optimization and updating stage, the system continuously improves the alternative indicators and evaluation models through self-learning mechanisms and continuous data accumulation. The system continuously collects new traffic data during operation and stores it in the data warehouse. The evaluation model is regularly evaluated and retrained to optimize model parameters and improve prediction performance.

[0244] According to the latest research results and application requirements, new alternative indicators are introduced and indicator weights are adjusted to ensure the scientific nature of the evaluation results. Through software updates, hardware maintenance and security protection, the efficiency and security of the system are maintained, while user feedback is collected to improve the system and optimize the user experience.

[0245] Through the above steps, the system can be continuously optimized and updated in practical applications, maintaining its leading position in road safety monitoring and evaluation, and continuously improving the level of road traffic safety.

[0246] Example 3

[0247] This embodiment provides an optimal traffic safety alternative indicator selection system applicable to complex road nodes, including:

[0248] Data acquisition and preprocessing module: used to define the boundaries of road areas at complex road nodes and store them in the database in the form of coordinates; relying on roadside sensing equipment, it periodically acquires the complete microscopic trajectory data of all vehicles in complex road nodes, and performs data preprocessing and parameter extraction;

[0249] Dangerous scene data calculation and acquisition module: used to determine and verify dangerous driving scenes at complex road nodes based on vehicle trajectory data samples and extracted parameters, and obtain dangerous scene data;

[0250] Candidate traffic safety alternative index value acquisition module: used to construct a set of candidate traffic safety alternative indexes, and calculate and obtain the value of each candidate traffic safety alternative index for dangerous scenarios;

[0251] The module for obtaining the optimal traffic safety alternative indicator is used to construct a set of evaluation indicators for traffic safety alternative indicators. For each candidate traffic safety alternative indicator, the module calculates and obtains the value of each evaluation indicator, where the evaluation indicators include accuracy, rapidity, and severity. According to the value of each evaluation indicator, the module calculates the distance between each candidate traffic safety alternative indicator and the corresponding optimal solution and the worst solution, and selects the traffic safety alternative indicator with the highest relative proximity to the optimal solution as the optimal traffic safety alternative indicator.

[0252] It can be understood that the same or similar parts of the above embodiments can be referenced to each other, and the contents not described in detail in some embodiments can refer to the same or similar contents in other embodiments.

[0253] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

[0254] It should be understood that in the embodiments of the present invention, the processor may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The memory may include a read-only memory and a random access memory, and provide instructions and data to the processor. A portion of the memory may also include a non-volatile random access memory. For example, the memory may also store information about the type of device.

[0255] The readable storage medium is a computer-readable storage medium, which may be an internal storage unit of the controller described in any of the foregoing embodiments, such as a hard disk or memory of the controller. The readable storage medium may also be an external storage device of the controller, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the controller. Furthermore, the readable storage medium may also include both an internal storage unit of the controller and an external storage device. The readable storage medium is used to store the computer program and other programs and data required by the controller. The readable storage medium may also be used to temporarily store data that has been output or is to be output.

[0256] Based on this understanding, the technical solution of the present invention, in essence or in other words, the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned readable storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., various media that can store program codes.

Claims

1. A method for selecting optimal traffic safety alternative indicators for complex road nodes. It is characterized in that include: S1: Delineate the boundaries of the road area of ​​complex road nodes and store them in the database in the form of coordinates; S2: Periodically obtain the microscopic trajectory data of all vehicles in the complex road nodes, perform data preprocessing to obtain vehicle trajectory data samples, and further extract parameters; S3: Based on the vehicle trajectory data samples and extracted parameters, the dangerous driving scenarios at complex road nodes are determined and verified to obtain dangerous scenario data; S4: Construct a set of candidate traffic safety alternative indicators, calculate the values ​​of each candidate traffic safety alternative indicator for dangerous scenarios, and divide them into low-risk and high-risk states; S5: construct an evaluation index set of traffic safety alternative indicators, and calculate and obtain the value of each evaluation index for each candidate traffic safety alternative indicator, wherein the evaluation index includes accuracy, rapidity, and severity; S6: According to the values ​​of each evaluation index, calculate the distance between each candidate traffic safety alternative index and the corresponding optimal solution and the worst solution, and select the traffic safety alternative index with the highest relative proximity to the optimal solution as the optimal traffic safety alternative index.

2. The optimal traffic safety alternative indicator selection method applicable to complex road nodes according to claim 1, It is characterized in that The specific process of S1 is: S11. Delineation of road area boundaries at complex road nodes: Identify complex road nodes using high-precision map data and satellite images; select multiple representative points within the identified area; and generate boundary polygons of the road area using a boundary fitting algorithm based on the selected representative points; S12. Establishing a database based on the acquired complex road node road area boundary coordinates: performing coordinate conversion on the coordinates of each vertex of the generated boundary polygon; storing the converted boundary coordinate data in a structured form in a database, wherein the database adopts a relational database or a spatial database; S13. Based on the established database, verify the boundary data stored in the database; and adjust and correct inaccurate or incomplete boundary data based on the verification results.

3. The optimal traffic safety alternative indicator selection method applicable to complex road nodes according to claim 1, It is characterized in that The S2 step includes: S21. The acquisition of vehicle micro-trajectory data at complex road nodes is to collect vehicle motion trajectory data in the entire area of ​​the complex road node using roadside sensing equipment or aerial photography, and perform data preprocessing, wherein the data preprocessing includes: Data cleaning: remove invalid values, error values ​​and duplicate values ​​in the data; Data synchronization: synchronize the time of data acquired by different sensors to ensure data consistency; Time alignment: Align the time series of data collected at different time points to ensure the consistency of the data on the time axis; Spatial correction: For the data obtained by aerial photography or roadside sensing equipment, spatial correction is required to convert the image data into actual ground coordinates to ensure the accuracy of the data; S22. Process the vehicle trajectory data samples through the roadside edge computing device, extract parameters, and obtain vehicle characteristic parameters and environmental characteristic parameters.

4. The optimal traffic safety alternative indicator selection method applicable to complex road nodes according to claim 1, It is characterized in that The determination of dangerous driving scenarios is based on the vehicle feature parameters extracted from the periodically collected vehicle trajectory data samples after preprocessing, and discriminates each vehicle C within the sample f The size relationship between the instantaneous monitoring distance d between the center point of the front of the vehicle and the center point of the rear of the preceding vehicle C l at each moment t and the braking distance S of the vehicle itself. When d < S, it is determined that the start time of the dangerous driving scenario of vehicle C f occurs When d is less than or equal to 0.5 meters, it is determined that a collision has occurred and the end time of the dangerous scenario A total of n dangerous driving scenario data are obtained, and this dangerous scenario data is the reference set A of traffic safety alternative indicators; Where d is the instantaneous monitoring distance, that is, vehicle C f The center point of the front vehicle and the front vehicle C l The spatial distance between the center points of the rear of the vehicle; S is the vehicle C f The braking distance is calculated as follows: Where, t 1 The time from when the driver starts to step on the brake pedal to when the brake starts to take effect; t 2 The time for the braking force to increase; v 0 is the initial braking speed of the vehicle; j max is the maximum braking deceleration; And according to each vehicle C f The center point of the front of the vehicle is at the same time as the front vehicle C at every moment t. l The instantaneous monitoring distance d between the center points of the rear of the vehicle divides all dangerous scenes into two states: low-risk state and high-risk state: Low-risk state, i.e., each vehicle C f is in the state of X ≤ d < S, where X is the safety braking distance of the leading vehicle C l when the leading vehicle is driving normally, and the following vehicle C f The calculation formula is as follows: v 1 For the front car C l speed. High-risk state, i.e., each vehicle C f is in a state where 0 ≤ d < X.

5. The optimal traffic safety alternative index selection method applicable to complex road nodes according to claim 4, It is characterized in that The S4 step includes: S41. Construct a set of candidate traffic safety alternative indicators, where the candidate traffic safety alternative indicators include: distance collision time TTC, improved distance collision time MTTC, extended distance collision time ETTC, interference collision time TTCD, collision avoidance deceleration rate DRAC, composite indicator COM ettc ; S42. For the dangerous scene reference set A, calculate and obtain the specific value of each candidate traffic safety alternative indicator for each frame; S43. Classify each candidate traffic safety alternative indicator into a state, including a low-risk state and a high-risk state: when the value of each indicator is within the threshold range of the preset low-risk state, it is classified as a low-risk state; when the value of each indicator is within the threshold range of the preset high-risk state, it is classified as a high-risk state; Among them, the composite index COM ettc It is based on a vehicle in a scene at a certain time period T = [t 0 -t * , t 0 +t * ] within the mean ETTC mean , extreme ETTC min and t 0 The instantaneous ETTC value at the time Constructed; where t 0 is any time in the dangerous scene reference set A, t * is the preset time interval length, COM ettc The value of is calculated once per frame; Calculate the instantaneous ETTC for each frame t t value, t is any time, and the instantaneous ETTC t Value preprocessing: When ETTC t <0s or ETTC t >10s, set ETTC t =10s,ETTC mean is the vehicle C in time period T f ETTC at each time t t The mean of ETTC min is the vehicle C in time period T f ETTC t The minimum value of t 0 The instantaneous ETTC value at the time; ETTC mean ETTC min The calculation formula is as follows:

6. The optimal traffic safety alternative index selection method applicable to complex road nodes according to claim 5, It is characterized in that Based on three basic indicators ETTC mean and ETTC min Construction of composite indicator COM ettc The specific process is: S421: All time periods T divided by the dangerous scene reference set A = [t 0 -t * , t 0 +t * ] to standardize the three basic indicators Where Y ij is the standardized value of the jth basic indicator in the i-th period; x ij is the value of the jth basic indicator in the i-th period, i = 1, 2, ..., n; j = 1, 2, 3, corresponding to the three basic indicators respectively ETTC mean and ETTC min ; S422. Based on the standardized value Y ij , calculate three basic indicators ETTC mean and ETTC min The information entropy is calculated as follows: In the formula, E j is the information entropy of the jth basic indicator; f ij Y is the proportion of the i-th period under the j-th basic indicator to the sum of all sample values ​​of the indicator; ij is the standardized value of the jth basic indicator in the i-th period, i = 1, 2, ..., n; j = 1, 2, 3, corresponding to the three basic indicators respectively ETTC mean and ETTC min ; S423. Based on three basic indicators ETTC mean and ETTC min Information entropy E j , calculate the difference coefficient and entropy weight of the three basic indicators: G j =1-E j In={In 1 ,IN 2 ,IN 3 } Where W j is the entropy weight under the jth basic index; G j is the difference coefficient of the jth basic index, j = 1, 2, 3, corresponding to the three basic indicators respectively ETTC mean and ETTC min ; W is the entropy weight matrix of the three basic indicators; S424. Based on the standardized value Y ij and entropy weight W j , construct a weighted decision matrix: Z ij =Y ij ×W j Among them, Z is the weighted decision matrix; Y ij is the standardized value of the jth basic indicator in the i-th period, Z ij W is the weighted normalized value of the jth basic indicator in the ith period; j is the entropy weight under the jth basic index; i = 1, 2, ..., n; j = 1, 2, 3, corresponding to the three basic indicators respectively ETTC mean and ETTC min ; S425. Based on the calculated weighted decision matrix Z, obtain the time period with the least conflict and the time period with the greatest conflict among all time periods; wherein the time period with the least conflict refers to the maximum value of each basic indicator in the weighted decision matrix; the time period with the greatest conflict refers to the minimum value of each evaluation indicator in the weighted decision matrix: Among them, Z + The period with the least conflict; Z - This is the time when the conflict is greatest; S426. Calculate the distance between each time period and the time period with the least conflict and the time period with the greatest conflict under the jth basic index: in, is the distance between the i-th time period and the time period with the least conflict; is the distance between the i-th time period and the time period with the largest conflict; is the period with the least conflict under the jth basic indicator; Z ij It is the normalized value after weighting under the jth basic indicator in the ith period; is the period with the largest conflict under the jth basic indicator; i = 1, 2, ..., n; j = 1, 2, 3, corresponding to the three basic indicators respectively ETTC mean and ETTC min ; S427. Based on the distance between each time period and the time period with the least conflict and the time period with the greatest conflict Calculate the relative proximity S between the i-th continuous conflict period and the period with the least conflict i , that is, the central time t of the i-th period 0 Composite index Value, where the relative closeness S i The calculation formula is as follows: in, is the distance between the i-th time period and the time period with the largest conflict; is the distance between the i-th time period and the time period with the least conflict; i = 1, 2, ..., n; S i is the relative proximity of the i-th time period to the time period with the least conflict.

7. The optimal traffic safety alternative index selection method applicable to complex road nodes according to claim 1, It is characterized in that The evaluation index set of the traffic safety alternative index in S5 refers to a set of evaluation indexes for evaluating the performance of all traffic safety alternative indexes, wherein the evaluation indexes include accuracy, rapidity, and severity: Accuracy refers to the accuracy of each traffic conflict alternative indicator in identifying dangerous scenes. Its calculation refers to the ratio of the number of scenes in which the current traffic conflict alternative indicator judges each scene as low-risk or high-risk to the number of all dangerous scenes among the n dangerous scenes in the selected reference set A. Rapidity refers to the rapidity of each traffic conflict alternative index in identifying dangerous or high-risk scenes; its calculation refers to the rapidity of identifying dangerous or high-risk scenes for any dangerous scene in the reference set A, starting at the i-th scene To determine the standard time, each traffic conflict alternative index x is used to judge the dangerous state of the same scene. The time when each index first determines the i-th scene as dangerous or high-risk is Get each indicator The time and the i-th scene The time difference between the moments When Δt i，x When it is positive, it means that the indicator judges the dangerous state of the scene earlier than the manual judgment in the reference set A. At this time, the rapidity of the judgment of each indicator can be expressed as and is a positive value in the range [0, 1], where It is the fastest indicator to determine the accuracy among all indicators. The time and the i-th scene The time difference between the moments; when Δt i，x When it is a negative value, it means that the indicator’s judgment of the dangerous state of the scene is later than the manual judgment in the reference set A. At this time, the rapidity of the judgment of each indicator can be expressed as in is the end time of the i-th scene, and It is a negative value, and its value range is (0, -∞); Severity refers to the accuracy of each traffic conflict alternative index in identifying various dangerous states of dangerous / high-risk scenes. Its calculation refers to the ratio of the number of dangerous scenes judged by the current traffic conflict alternative index to all dangerous scenes in the n dangerous scenes in the selected reference set A. 1 , and the proportion of the number of high-risk scenes in each scene to the number of all high-risk scenes P 2 , select P 1 and P 2 The average value is the severity evaluation index.

8. The optimal traffic safety alternative index selection method applicable to complex road nodes according to claim 1, It is characterized in that The specific process of S6 is as follows: S61: Standardize the accuracy, rapidity, and severity values ​​of all candidate traffic safety alternative indicators: Among them, Y ij is the standardized value of the i-th safety alternative indicator under the j-th evaluation indicator; x ij is the value of the i-th safety alternative index under the j-th evaluation index; i is the safety alternative index, i=1, 2, ..., n; j is the evaluation index, when j=1 it is accuracy, when j=2 it is rapidity, when j=3 it is severity; S62: The standardized value Y of each safety alternative indicator under each evaluation indicator ij , calculate the information entropy E of each evaluation index j and the difference coefficient G j , the formula is as follows: G j =1-E j In the formula, E j is the information entropy of the jth evaluation index, when j = 1 it is accuracy, when j = 2 it is rapidity, when j = 3 it is severity; f ij Y is the proportion of the i-th safety alternative indicator under the j-th evaluation indicator to all sample values ​​of the indicator, i = 1, 2, ..., n; ij is the standardized value of the i-th safety alternative indicator under the j-th evaluation indicator; G j is the difference coefficient of the jth evaluation index; S63: Based on the obtained difference coefficient of the j-th evaluation index, calculate and obtain the entropy weight W corresponding to the j-th evaluation index j And establish the entropy weight matrix W: In={In 1 ,IN 2 ,IN 3 } Among them, W j is the entropy weight under the jth evaluation index, when j = 1 it is accuracy, when j = 2 it is rapidity, when j = 3 it is severity; G j is the difference coefficient of the jth evaluation index; W is the entropy weight matrix; S64: Based on the normalized value Y ij and entropy weight W j , construct a weighted decision matrix: Z ij =Y ij ×W j Among them, Z is the weighted decision matrix; Y ij is the standardized value of the i-th safety alternative indicator under the j-th evaluation indicator, i = 1, 2, ..., n, when j = 1, it is accuracy, when j = 2, it is rapidity, when j = 3, it is severity; W j is the entropy weight under the jth evaluation index; S65: The maximum value of each evaluation index in the weighted decision matrix Z is defined as the optimal solution; the minimum value of each evaluation index in the weighted decision matrix Z is defined as the worst solution: Among them, Z + is the optimal solution; Z - is the worst solution; S66: The optimal solution Z based on each evaluation indicator + and the worst solution Z - , calculate the distance between each candidate traffic safety alternative index and the optimal solution and the worst solution under the jth evaluation index: in, is the distance between the ith traffic safety alternative index and the optimal solution; is the distance between the ith traffic safety alternative index and the optimal solution; is the optimal solution under the jth evaluation index; Z ij is the weighted value of the i-th safety alternative indicator under the j-th evaluation indicator; is the worst solution under the jth evaluation index, i = 1, 2, ..., n, when j = 1 it is accuracy, when j = 2 it is speed, when j = 3 it is severity; S67: Based on the distance between each traffic safety alternative index and the optimal and worst solutions Calculate the relative closeness S between the i-th safety alternative index and the optimal solution i , where the relative proximity S i The calculation formula is as follows: Among them, S i is the relative closeness between the ith traffic safety alternative index and the optimal solution; is the distance between the ith traffic safety alternative index and the worst solution; is the distance between the ith traffic safety alternative index and the optimal solution, i = 1, 2, ..., n, S68: Select relative proximity S i The highest safety alternative is the most optimal traffic safety alternative.

9. An optimal traffic safety alternative index selection system suitable for complex road nodes. It is characterized in that include: Data acquisition and preprocessing module; It is used to define the boundaries of the road area of ​​complex road nodes and store them in the database in the form of coordinates; it periodically obtains the complete microscopic trajectory data of all vehicles in the complex road nodes, and performs data preprocessing to obtain vehicle trajectory data samples, and further extracts parameters; Dangerous scene data calculation and acquisition module: used to determine and verify dangerous driving scenes at complex road nodes based on vehicle trajectory data samples and extracted parameters, and obtain dangerous scene data; Candidate traffic safety alternative index value acquisition module: used to construct a set of candidate traffic safety alternative indexes, and calculate and obtain the value of each candidate traffic safety alternative index for dangerous scenarios; The module for obtaining the optimal traffic safety alternative indicator is used to construct a set of evaluation indicators for traffic safety alternative indicators. For each candidate traffic safety alternative indicator, the module calculates and obtains the value of each evaluation indicator, where the evaluation indicators include accuracy, rapidity, and severity. According to the value of each evaluation indicator, the module calculates the distance between each candidate traffic safety alternative indicator and the corresponding optimal solution and the worst solution, and selects the traffic safety alternative indicator with the highest relative proximity to the optimal solution as the optimal traffic safety alternative indicator.

Images